This is long, but I put lots of purely human effort* into it, and I hope it clears some things up. Writing it cleared up a lot in my own head.

> I would much rather read a human comment with typos and poor grammar than another piece of anodyne LLM output that shows only that the responsible party doesn't value the human interaction that I do.

I take your meaning. However, that phrasing doesn't cut to the core of it. Rationalists would say "this doesn't carve reality at the joints". Here are my attempt to disentangle, decomplect**, and find common ground. Let me know which of these you disagree with, if any:

1. I care relatively little about typos and grammar, as long as the ideas are clear.

2. I enjoy human connection with people, in person and online. I would prefer to have a person on the other end.

3. Actually, I'd go further... I'd like to have more personable conversations and work past a lot of the common online discussion failure modes (but now I'm wandering off topic).

4. When chatting, I care a lot about the quality of the underlying thinking.

5. I personally don't want to read someone's first knee-jerk take.

6. I prefer to read a thoughtful and clear expression of a human being's experience.

7. On HN in general, I want curious conversation.

8. I understand everybody brings some point of view and sometimes what one would call an agenda.

9. Maybe the top criteria for doing #6 well is: are we interacting with each other per the guidelines? Charitably, in good faith, and with curiosity (#7).

10. As an example of an unwelcome agenda (#8), I don't want to inundated with commercially-fueled marketing-speak. However, to be clear, in this regard, I don't care if it comes from a person or an LLM.

11. Speaking for myself, not for HN, I don't mind if #6 is assisted by LLM editing.

12. Why #11? Because I care more about having a human being in the loop than a human shaping every single aspect. (See also the next point.)

13. One key for me is: does the person stand behind what they post? Meaning: are they accountable to it? Do they own it?

14. In addition to "original thoughts" (as if humans ever really have them!), #13 applies to someone borrowing, remixing, or outright stealing phrases they've heard before.

15. If someone uses an LLM to edit their words, it feels not too different from #14. Except when ... (see #16)

16. Sometimes people use LLMs to not actually put in the work of reflecting and thinking. This is sad for them and sad for people who have to read it.

17. Unfortunately, even without LLMs, some people don't put in the work of thinking and reflecting. See #5.

18. Putting #17 and #18 together, it is NOT the part about the LLM that bothers me! It is the lack of reflecting and thinking!

19. Asking someone else to read your post before sending is totally ok.

20. #15 done well may not be different (and may be better than!) #19.

21. In a forum where a comment is read many more times than it is written, I consider it more respectful to put an appropriate amount of effort into writing.

22. If a person takes the time to write something out in their own words, that is a signal of respect for the audience. Especially in comparison with, say, just replying with a trope.

23. If a person uses an LLM to research and clarify their thinking AND is thoughtful about it (#6) AND stands behind it (#13), that is a signal for respect for the audience.

Fin.***

Here's what I'm driving at: I recommend that people put the effort in to figure out what aspects bother them about this moment in time with so much GenAI output. It is a real PITA to make sense of, but not doing so doesn't make that PITA go away.

* I hope you don't mind that I used a NCLM, a neuro-cognitive language model, to construct this... a.k.a my brain. Snark aside, does the substrate matter, in the long-run?

** Rich Hickey is my home boy.

*** Why do I number my points? Well, I work hard to tease apart my ideas; it is part of my writing and thinking process. Sometimes I put them back into a paragraph, sometimes not. But I like trying it out: I think it makes it easier to refer back to ideas and build upon them.